Advanced Materials Research Vols. 532-533

Abstract: In order to solve the problems of obtaining users’ individual demand in the current field of adaptive mobile map service, this paper proposes solutions of adaptive mobile map based on spatiotemporal data mining and architecture; and discusses several key technologies. The study of adaptive mobile map based on spatiotemporal data mining can improve adaptive mobile map service, and implement obtain efficiently useful information from a massive data to meet the diverse needs of different users.

Abstract: Low-Density Parity-Check(LDPC) codes are a class of channel codes based on matrix encoding and iterative decoding. It has low decoding complexity as well as capacity approaching performance. Until now, the best designed LDPC codes can achieve the performance within only 0.0045dB of the Shannon limit. With the in-depth study, the encoding complexity of LDPC codes is not a difficult problem for application any more. Today, we can see LDPC codes widely used in many practical systems, such as wireless communication system, deep-space communication system, optical-fiber communication system and media storage system. This thesis first introduces the development of channel coding, and then the basic principles and concepts of LDPC codes. The following parts discuss several techniques of LDPC codes, including the construction methods of low-density parity matrix, the iterative decoding algorithms and performance analysis methods. Besides, we propose our opinions and our improved algorithms.

Abstract: A data-centric approach to business process and workflow modeling has been emerging over the past several years. But various research interests of access contral are still faced with quite a number of challenges. In this paper,we discuss the key access control requirements for application data in business process applications by using employee hiring examples. As to achieve those requirements, we introduce a fine-grained access contral technology (FGAC) for the data-centric business process model.

Abstract: This paper discusses the methods of improving the efficiency and exactitude of the road conditions test though the modeling test algorithm. It presents the data structure of road model in which double direction data chain can be established. There are two clipping windows, W_out and W_contour. W_out would eliminate the lines that can’t intersect rapidly. The lines that maybe intersect would be estimated exactly by W_out after coordinates transforming.Thealgorithms of the road conditions testing are proved to be efficacious.

Abstract: In order to store real-time images, a digital image grabbing system is proposed in this paper. The principles and design of real-time image continuous grabbing, displaying and storing system are presented. It has been redeveloped based on Matrox Meteor-II /Digital Image Grabbing Board by the use of VC++, and the whole implementation was introduced by an example. The developed image capture software is given at the end of this paper. And this gives a new thought and platform for real-time image processing.

Abstract: The goal of image compression is to represent an image with as few number of bits as possible while keeping the quality of the original image. With the characteristics of higher compression ratio, fractal image coding has received much attention recently. However, conventional fractal compression approach needs more time to code the original image. In order to overcome the time-consuming issue, a Quadtree-based partitioning and matching scheme is proposed. During the partitioning phase, an image frame is partitioned into tree-structural segments. And during a matching phase, a rang block only searches its corresponding domain block around previous matched domain block. Such local matching procedures will not stop until a predefined matching threshold is obtained. The preliminary experimental results show that such sub-matching rather than a global matching scheme dramatically decreases the matching complexity, while preserving the quality of an approximate image to the original after decoding process. In particular, the proposed scheme improves the coding process up to 2 times against the conventional fractal image coding approach.

Abstract: Feature extraction plays an important role in speech recognition. In this paper, we propose a speech feature extraction scheme which focuses on increasing the robustness of speech recognizer in noise (additive) and channel (convolutive) distortion environment. Considering the two distortions are additive in spectral and log-spectral domain, respectively, we remove the additive components by computing the time derivatives of speech frames firstly in spectral domain and then in log-spectral domain. Compared with conventional methods, this method does not need spectrum estimation and prior knowledge of noise. Experimental results confirm that our proposed method can improve the speech recognition performance in environ-ments existing both noise and channel distortions.

Abstract: With the ever-increasing expansion of wireless voice and data plan demands, bandwidth constrained wireless channels intervened with mobility issues lay out a great challenge in the hardware configuration, software development and algorithm design of communication systems. This paper presents our work on a modular and standards-independent DSP-based testbed designed for real-time video communication over low-bandwidth channels. In particular, the real-time transmission of MPEG-4 video and JPEG2000 still images over the Bluetooth wireless standard is demonstrated here. The testbed was designed to explore the possibilities, demands, and challenges of real-time wireless video transmission. The work could have tremendous near-future applications.

Abstract: In order to raise speed of reading data from disk array memory, scientific and technological personnel have introduced cache technology into disk array. Since this technique has been invented, although the efficiency of reading data have been solved, after writing operation of countless times in disk cache, data consistency problem has been prominent expression. Especially in this condition that false of electricity and machine abnormal failure, the consistency of the data is more difficult to guarantee. In this paper, we adopt Non-Volatile RAM (NVRAM) devices to realize that the data will not be lost in disk array cache after power failure. Here we design a kind of new cache organizational structure. We firstly introduce cache structure of two tables (real-time mapping table, backup mapping table) and a cache backup block. Then through these two tables, we can work to recover data through the copy between the two tables in the macroscopic, and in the microscopic through cache backup block can backup the cache data from writing failure. As power failure and system breakdown, we can ensure that data will not easily lose and the original data can recovery after system crash by this technology. Thus it ensures the consistency of the data cache.

Abstract: The current study based on XML index and query mostly focuses on encoding and the structural relation. Region codings are widely used to improve XML query. In this paper postorder-traversal region coding is proposed. The postorder of a node’s all descendants consists of the region. Judging and ensuring structural relation of any two nodes just depend on this region, if the postorder of a node is in a region, ancestor/descendant structural relation can be ensured. Consequently, postorder-traversal region coding can effectively judge structural relation and avoid traversing the XML document tree. Based on region coding, many constructive structural query algorithms have been put forward. As we all know that Stack-Tree-Desc algorithm is one of these fine algorithms, AList and DList only need separately scan one time to judge structural relation, however some unnecessary nodes still be scanned. In order to solve this problem, Indexed Sequential Table algorithm is introduced. The optimized algorithm introduces Indexed Sequential Table to avoid scanning unwanted nodes when the two lists join to locate next node which participates in structural join. In this case, some nodes of AList and DList which don’t participate in structural joins can be jumped, the query efficiency is enhanced. As a result, ordered scanning is prevented, the consuming time of XML query shortens accordingly. Experiment results demonstrate the effectiveness of the improved coding and algorithm.